A pH-based biosensor for detection of arsenic in drinking water

A pH-based biosensor for detection of arsenic in drinking water
of 9
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  ORIGINAL PAPER  A pH-based biosensor for detection of arsenic in drinkingwater K. de Mora  &  N. Joshi  &  B. L. Balint  &  F. B. Ward  & A. Elfick   &  C. E. French Received: 13 October 2010 /Revised: 13 February 2011 /Accepted: 16 February 2011 /Published online: 27 March 2011 # Springer-Verlag 2011 Abstract  Arsenic contaminated groundwater is estimatedto affect over 100 million people worldwide, with Bangla-desh and West Bengal being among the worst affectedregions. A simple, cheap, accurate and disposable device isrequired for arsenic field testing. We have previouslydescribed a novel biosensor for arsenic in which the output is a change in pH, which can be detected visually as a colour change by the use of a pH indicator. Here, we present animproved formulation allowing sensitive and accuratedetection of less than 10 ppb arsenate with static overnight incubation. Furthermore, we describe a cheap and simplehigh-throughput system for simultaneous monitoring of pHin multiple assays over time. Up to 50 samples can bemonitored continuously over the desired time period. Cellscan be stored and distributed in either air-dried or freeze-dried form. This system was successfully tested on arsenic-contaminated groundwater samples from the South East region of Hungary. We hope to continue to develop thissensor to produce a device suitable for field trials. Keywords  Biosensors.Groundwater .Arsenic.Assay.BioBricks Introduction Arsenic in drinking water, in the form of arsenate or arsenite anions, is a major public health issue in a number of regions worldwide and is prevalent in South and SouthEast Asia, especially in Bangladesh and West Bengal [1, 2]. Recent studies have shown that arsenic poisoning may beworse than previously thought, with as many as 77 million people possibly exposed to arsenic contaminated water inBangladesh alone [3]. Studies have also shown groundwa-ter contaminated with arsenic at concentrations up to500 ppb in one of India ’ s most heavily industrialisedregions: the Thane region [4]. Arsenic is immediately toxicin high concentrations ( ∼ 60,000 ppb), but exhibits chroniceffects at lower concentrations [5]. The World HealthOrganization has set a maximum safe limit of 10 ppb(10  μ  g l − 1 ), but many regions still consider 50 ppb(50  μ  g l − 1 ) to be a safe target at the present time. Symptomsof inorganic arsenic poisoning can occur at concentrationsvarying from 300  –  30,000 ppb and include vomiting,stomach and intestinal irritation, nausea and diarrhoea [5].One of the most visible signs of arsenic poisoning is skinlesions that can appear on the hands, feet and torso. Theselesions can eventually result in skin cancer if untreated, but more often lead to social ostracism as arsenicosis is perceived Published in the special issue  Microorganisms for Analysis  with Guest Editor Gérald Thouand.K. de Mora :  A. Elfick School of Engineering, University of Edinburgh,Edinburgh EH9 3JL, UK  N. JoshiSchool of GeoSciences, University of Edinburgh,Edinburgh EH9 3JW, UK B. L. Balint Department of Biochemistry and Molecular Biology,University of Debrecen Medical and Health Science Centre,Debrecen H-4032, HungaryF. B. Ward :  C. E. FrenchSchool of Biological Sciences, University of Edinburgh,Edinburgh EH9 3JR, UK C. E. French ( * )University of Edinburgh,Darwin Building, King ’ s Buildings,Edinburgh EH9 3JR, UK e-mail: Anal Bioanal Chem (2011) 400:1031  –  1039DOI 10.1007/s00216-011-4815-8  to be a contagious disease [6]. At lower concentrations,visible symptoms can take months or years to appear.Current testing in the affected regions is most widelyconducted using atomic adsorption spectroscopy (M.Owens, personal communication). This technique is quan-titative and reliable [7] but problematic, as water samplesmust be shipped from contaminated sites to testinglaboratories. The capital investment per system is in theorder of 20,000 USD, with an additional cost for consum-ables and the use of a qualified technician. Various methodsalso exist to test for arsenic in the field, most notably theWagtech Arsenator and paper test strips based on theGutzeit method. The Arsenator produces a quantitativereadout of the arsenic concentration based on an electronicread of the Gutzeit method [8]. The basic mechanism behindthe Gutzeit-based field test strips is conversion of arseniccompounds into arsenic trihydride by zinc. The hydride thenstains paper impregnated with mercuric bromide [9].Although assays based on the Gutzeit method have beendeployed in the past, researchers are looking towardsenzymatic and biological approaches to detect arsenic.One approach is to use screen-printed electrodes (SPCE)with acetylcholinesterase (Ach) and to measure an amper-ometric response. Sanllorente-Mendez et al. have demon-strated an approach where they can detect arsenic to a limit of 1.1×10 − 8 M for their Ach/SPCE biosensor [10].Although successful, they reported storing their electrodesat 4 °C, which may make this device difficult to transport and store in environments lacking refrigeration. Other groups have reported using the electrostatic oxidation of  L -cysteine on screen-printed electrodes to determine arse-nate concentration. Sarkar et al. immobilised  L -cysteine ona working electrode by in situ polymerization of acrylamideand determined that their system produced a linear responseto arsenic from below 1 to 30 ppb [11]. This work showedfunctional and sensitive arsenic detection using screen- printed electrodes; however, the most sensitive platinumvariants may not be cost-efficient in terms of developing afield test device.Whole-cell biosensors offer a potential alternativemethod for arsenic detection [12]. Previous arsenic bioas-say systems have been developed using bioluminescence,Green Fluorescent Protein, and  β -galactosidase as thereporter mechanisms [13, 14]. Previous work from van der  Meerandcolleaguesdemonstratedasystemcombiningwholecell  Escherichia coli  biosensors with a microfluidics envi-ronment allowing arsenic concentration measurements to betaken from as few as 200 cells [15]. Bacteriological arsenic biosensors have been also been previously field tested [16].We have previously described a biosensor based ongenetically modified  E  .  coli  cells in which the presence of arsenate or arsenite induces expression of   β -galactosidase,allowing fermentation of lactose and consequent change in pH which can be read using a simple pH indicator [17, 18]. Here, we describe further characterization of the strain, animproved formulation of assay medium that allows for staticincubation of the samples and a new assay allowingcontinuous monitoring of up to 50 samples simultaneously. Materials and methods Organisms and growth conditionsConstruction of our arsenic biosensor organism,  E. coli JM109/pSB1A2-BBa_J33203, has been previously de-scribed [17]. The BioBrick assay construct, pSB1A2-BBa_J33203, is available from the Registry of StandardBiological Parts, hosted at Massachusetts Institute of Technology, as are the subcomponents BBa_J33201 (arse-nic-responsive promoter) and BBa_J33202 ( lacZ  ’   reporter gene) used to generate it. Initial experiments were con-ducted in Luria-Bertani (LB) broth with 10 g/l lactose and100  μ  g/ml ampicillin. Later experiments used a variant of Hugh and Leifson ’ s OF (oxidation  –  fermentation) mediumdesignated Arsenic Biosensor Medium (ABM) 6. ABM6contains 2.0 peptone, 0.1 yeast extract, 0.3 K  2 HPO 4 , 2.1 NaHCO 3 , and 0.1 g/l bromothymol blue. Lactose (10 g/l)was added if required, but in most cases was present in thedried cells used as inoculum as described below. Allcomponents were mixed prior to autoclaving. Arsenic wasadded as sodium arsenate from sterilised stock solutions at 1,000 or 10,000 ppb arsenic. Variants of ABM6 media weremade with 1,500 mg l − 1 of iron (II) sulphate (527 mg l − 1 iron) and 2 mg l − 1 zinc sulphate (0.76 mg l − 1 zinc)concentrations, respectively, representing nine and four times the highest reported concentrations in the BritishGeological Survey of Bangladeshi groundwater [19]. For  preparation of air-dried cells, overnight cultures weregrown in LB, mixed with an equal volume of 20%  w/v  lactose, dispensed at the appropriate volume into 1.5 mlmicrocentrifuge tubes, and incubated with open lids at 37 °Covernight, during which the cell  –  lactose mixture dried toa glassy state. For freeze-drying, overnight cultures wereharvested by centrifugation, resuspended in 20%  w/v   sterilelactose solution, dispensed into 1.5 ml microcentrifugetubes, frozen at   − 80 °C and dried in an Edwards Modulyofreeze drier for 15 h. Dried cells were stored at roomtemperature.Assay proceduresIn initial experiments, assays were conducted as described previously [17, 18], with 6 ml LB plus lactose and other  additives in 20 ml sealed glass vials, incubated withshaking at 37 °C, and pH was measured at intervals using 1032 K. de Mora et al.  a standard semi-micro glass pH electrode. In later assays, asspecified, 1 ml ABM6 was added to a tube of dried cells,and the desired concentration of sodium arsenate was addedfrom a stock solution. Unless otherwise specified, tubeswere then incubated at 37 °C without shaking. For simultaneous visual monitoring of multiple assays, up to50 tubes were placed in a specially constructed rack in a37 °C incubator with a clear front cover. Images werecaptured using a Creative Live! Vista IM webcam (modelno: VF0260) and Nimisis Flix time-lapse software was usedfor image acquisition. Jpeg images were acquired due to thesmall size of the files and the .AVI time-lapse videos weremade with no file compression in Flix. Due to file header and footer issues, the .AVI videos were opened and re-saved in ImageJ before being processed in Image Pro 7.0.Once parsed in Image Pro 7.0, data were processed intoExcel readable spreadsheets using Mathworks Matlab.Miller assays for   β -galactosidase activity were con-ducted according to a standard protocol [20]. Activity of XylE (catechol-2,3-dioxygenase) was monitored by addingcatechol to a final concentration of 0.1 mM to cell suspensionand monitoring absorbance of the yellow reaction product (2-hydroxy- cis , cis -muconic semialdehyde) at 420 nm. Results Effect of medium and inoculum levelPreviously, we have described assays in continuouslyshaken LB cultures inoculated from a previous liquidculture, with pH monitored at intervals using a pH electrode[18]. Clearly, this assay configuration is not ideal for fielduse. We therefore sought to develop a system with moresuitable characteristics for use by non-experts under fieldconditions with minimal laboratory equipment. To clarifythe effect of different growth parameters, initially, weinvestigated the effect of using a more dilute growthmedium. Use of LB diluted by a factor of 2 or 4 showedthe higher dilution factor medium led to less pH decrease inarsenic-free controls while maintaining a strong pHresponse to 60 ppb arsenic, but further dilution led to adecreased pH response in the presence of arsenic. Use of different levels of inoculum was also investigated. A larger inoculum led to more rapid response, but where sampleswere incubated for longer periods, the level of initialinoculum made little difference to the final pH over therange tested (data not shown).Effect of bicarbonateIn our previous characterization tests, we sought todetermine the effect of common groundwater buffer ionson pH response [18]. Phosphate at relevant levels wasfound to have no effect on the assay but bicarbonate wasunexpectedly found to apparently increase the sensitivity of the assay, leading to enhanced pH response at low arsenicconditions. To determine the basis for this effect, weconstructed a modified version of the biosensor by addinga second reporter gene,  xylE   (encoding catechol-2,3-dioxygenase; BioBrick BBa_J33204) to the BioBrick  biosensor plasmid, giving the new biosensor construct  pSB1A2-BBa_J15501. Biosensor organisms were incubat-ed in the presence of differing levels of bicarbonate andarsenate and were assayed for pH,  β -galactosidase activity(Miller assay) and XylE activity. As previously observed,the pH assay showed an enhanced response to low arsenateconcentrations in the presence of bicarbonate, whereas nodifference was seen at higher arsenate concentrations(Fig. 1a); thus, the presence of bicarbonate appeared toincrease the sensitivity of the assay. However, Miller assaysshowed that   β -galactosidase activity was increased in the presence of bicarbonate at both low and high arsenateconcentrations (Fig. 1b); thus, it appears that the apparent lack of effect of bicarbonate on the pH-based assay at higharsenate concentrations is an artefact caused by saturationof the pH response above a certain  β -galactosidase activity.Furthermore, despite a high background activity, XylEactivity was also seen to increase in the presence of arsenate(Fig. 1c), suggesting that the effect was due to increasedinduction of the reporter genes rather than some direct effect on the activity of   β -galactosidase. It therefore seemslikely that bicarbonate is increasing the interaction of arsenate with the ArsR repressor protein, perhaps byincreasing entry of arsenate into the cell or modifyingarsenate speciation. Further experiments may clarify this.For practical purposes, our previously described experi-ments [18] have indicated that the effect of bicarbonateappears to saturate at a concentration below 10 mM bicarbonate; thus, addition of sodium bicarbonate to allassays at a concentration above this should both enhancesensitivity and reduce variability due to varying bicarbonateconcentrations in water samples.Development of an assay medium suitable for staticincubationThe effect of different incubation conditions on assays inLB was also investigated, using a variety of different containers and liquid  –  air ratios (data not shown). The most significant result was that in unshaken cultures, a large pHdrop always occurred in arsenic-free controls, presumablydue to the decreased aeration, making the test unreliable.Since we aim to develop an assay suitable for field use, thisled us to investigate alternative formulations of assaymedium. The OF (oxidation  –  fermentation) test is a standard A pH-based biosensor for detection of arsenic in drinking water 1033  and widely used test for the production of acid fromcarbohydrates and is commonly used in the identification of medically important bacteria [21]. This assay is normallyconducted without shaking and uses the pH indicator  bromothymol blue to generate a visual response. Variousmodifications of OF medium, with the addition of bicar- bonate, were tested, eventually resulting in the development of ABM6 (see Materials and methods for composition),which was found to give a good colour response in lessthan 24 h in the presence of arsenate concentrations as lowas 5 ppb arsenic but not to change colour over 24 h in theabsence of arsenic, provided that a very low inoculum levelwas used (data not shown). This medium was used as the basis for further experiments.Drying of cells for storage and distributionDevelopment of this assay for field use will require that cells be dried for storage and distribution. We have previously reported that lactose is a suitable cryoprotectant for freeze-drying of cells, resulting in as high as 30%survival under the conditions tested. This also provideslactose required for the assay. Since our previous experi-ments had shown that a small inoculum provides the best results for overnight assays under unshaken conditions, weinvestigated the possibility of air-drying as an alternative tofreeze-drying, as this would remove the requirement for anexpensive piece of equipment. Cells (initial volume of 20 to100  μ  l per 1.5 ml microcentrifuge tube) were air-dried at 37 °C from a suspension containing 10%  w/v   lactose andwere found to dry overnight to a glassy consistency.Figure 2 shows survival of cells dried from initial volumesof 20 or 100  μ  l of cell suspension (Fig. 2). Thus, the tubescontaining 100  μ  l of cells initially showed higher survival but later showed more rapid die-off. This may be simplydue to less effective drying leading to higher moisturecontent when sealed. Nevertheless, tubes of biosensor cellsdried in this way showed reliable response to arsenic for at least 2 weeks after drying. Unless this can be improvedwith further development, freeze-drying will probably berequired for longer-term storage.Development of a colorimetric assayfor the high-throughput quantification of pHmeasurementsTofacilitatecharacterizationofourbiosensorsystem,we havedeveloped a technique to measure the pH of multiple samplessimultaneously over any desired period with high temporalresolution (Materials and methods). In order to calibrate theimage acquisition system, we measured the absorbance of water samples with bromothymol blue adjusted to pH valuesin the range 10.91 to 3.24 (Fig. 3). The greatest change inabsorbance was found to occur at 625 nm.In the multi-tube system, assays are conducted byinoculating a series of 1.5 ml microcentrifuge tubes withABM6 medium supplemented with arsenate, then placingthem on a rack in a 37 o C incubator in front of a Webcamera. Images are acquired over the period of theexperiment and compiled into a time lapse video which isread into Image Pro 7.0 and analysed using a custom-designed script. As the colour of the blue tubes stands out  45678123456   p   H   A  s  s  a  y   (  p   H   ) 020040060080010001200123456         β   -  g  a   l  a  c   t  o  s   i   d  a  s  e   (   M   i   l   l  e  r  u  n   i   t  s   )    X  y   I   E   A  s  s  a  y   (   A    4   2   0    ) a)b)c) Fig. 1  Effect of bicarbonate on reporter activity.  a  Initial andovernight pH,  b  β -galactosidase (Miller assay),  c  XyIE (catechol-2,3-dioxygenase) assay. Sample identification:  1  no bicarbonate, 0 ppbarsenic as arsenate.  2  no bicarbonate, 7.5 ppb arsenic as arsenate.  3  no bicarbonate, 15 ppb arsenic as arsenate.  4  12.5 mM bicarbonate, 0 ppbarsenic as arsenate.  5  12.5 mM bicarbonate, 7.5 ppb arsenic asarsenate.  6   12.5 mM bicarbonate, 15 ppb arsenic as arsenate1034 K. de Mora et al.  relative to the background, a colour threshold is applied tocreate regions of interest. For each region of interest, the red,green and blue colour channel intensity levels are averagedoverthe areaofthetubeandrecordedinanExcelspreadsheet.The scriptapplies the sameregionsofinteresttoeachframe inthe time-lapsecreatingRGB numerical valuesfor eachtubeat each time step. The Excel file is read into MathWorks Matlabto parse the data into a three-dimensional matrix of the red,green and blue colour intensities. Initial examination of theraw data confirmed that the greatest colour change occurredon the red channel as seen at 625 nm in Fig. 3. Plots showingthe colour change over time with bromothymol bluetherefore used data from the red channel.Calibration of the pH scale in the plot was achieved byimaging pH-adjusted samples of bromothymol blue. Thesuppliers of bromothymol blue (Sigma) specify a colour change over the pH range of 7.6 (blue) to 6 (yellow) withshades of green in the transition region. Figure 4 shows a plot of illumination against pH for the same pH-adjustedsamples used in Fig. 3. These data confirm that we canmeasure the colour change in samples over the 7.6  –  6 pHrange specified by the supplier.Once the assay configuration had been established andcalibrated, the first trial of the system was performed usingair-dried cells (dried from 50  μ  l of liquid consisting of 25  μ  l overnight culture plus 50  μ  l 20%  w/v   lactose; seeMaterials and methods above) following the addition of 1 ml ABM6 amended with various concentrations of arsenic (as sodium arsenate). Figure 5 shows the colour development over 60 h. The tubes are arranged in groups of three replicates in increasing concentration on the  X   axisand time on the  Y   axis. The general trend in the imageshows that the higher concentration samples change from blue to yellow more rapidly than the samples with lower arsenic concentration. Note that one tube in the 10 and the50 ppb arsenic concentrations did not change colour,indicating that further attention must be paid to qualitycontrol during preparation of assays.In Fig. 5, colour changes can be observed over time. Theexact moment of this change can be plotted if a quantitativemeasure of the colour of the tube can be obtained. Wemeasured the tube colour using the quantitative method previously described in the Materials and methods section.Figure 6 shows the quantified change in colour of the pHsamples over time. The data were processed by first normalising the cell-free controls to a pH of 7.6 as thesetubes do not change colour and are therefore suitable for use as a datum and for colour correction. The tubes areadjusted to the same starting pH value, which correspondsto a pH of 7.6 or above. As the experiment is conducted    A   b  s  o  r   b  a  n  c  e pH Camera illumination value Fig. 4  Numerical illumination value against pH for a range of pH-adjusted samples. Each data point represents a software processedillumination value for a pH calibrated 1 ml tube of water and bromothymol blue    A   b  s  o  r   b  a  n  c  e Wavelength (nm) BlankpH 10.91pH 9.21pH 7.60pH 7.35pH 7.96pH 6.75pH 6.35pH 5.96pH 4.54pH 3.24ABM6 Fig. 3  Absorbance spectrum of bromothymol blue at varying pHvalues 1.00E+031.00E+041.00E+051.00E+061.00E+071.00E+081.00E+09010203040    C   F   U   /  m   l Time (days) 20 µl cell suspension100 µl cell suspension Fig. 2  Survival of air-dried cells following storage at roomtemperatureA pH-based biosensor for detection of arsenic in drinking water 1035
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks